model training matrix
#1
by
TahirC
- opened
I’m trying to figure out how big the training dataset usually needs to be and how long it takes to train a LoRA for Qwen-Image-Edit.
I want to train one for personal use with the ostris AI Toolkit, but I can’t find any specifics for Qwen-Image-Edit anywhere. Just trying to estimate the time and compute budget I’d need.
Hey @TahirC
Note: This is not a training datasets. These are just test sample inferences of the adapter.
If you want to learn more about the Ostris AI Toolkit, please visit the channel directly.
For example, refer to this content:
Train a Qwen Image Edit 2509 LoRA with AI Toolkit - Under 10GB VRAM
prithivMLmods
changed discussion status to
closed